YouTube Monetization Half-true — works only if you do the unspoken work

By the Income Reality Check Editorial Team · Updated 2026-08-28 · 7 min read

YouTube Shorts built by Claude and ChatGPT: the $39K math the video skips

Verdict: Half-true — works only if you do the unspoken work. The revenue figures are plausible for a handful of channels, but the RPM the video uses is at the top of the range and the model it teaches is the kind YouTube can demonetize.

Sanji Nai-Chien’s video, “How a Faceless Channel Made $39K With Claude + ChatGPT,” had about 75,458 views when we reviewed it. The pitch: three faceless Shorts channels pulled in roughly $40,000, $25,000, and $33,000 over 90 days using nothing but Claude, ChatGPT, and an AI video tool called RankReel. Is that real? Partly. The numbers can happen. Getting there yourself is a different story than the one the video tells.

What the video actually claims

The creator says he’s been tracking three faceless “interesting facts” Shorts channels through VidIQ’s estimates. One did an estimated $40,000 from 250 million views in 90 days; another around $25,000; a third about $33,000. One channel, he adds, cleared two billion views and an estimated $170K over 365 days.

The workflow is genuinely simple to describe. ChatGPT generates the channel name, profile picture, banner, and description. Claude — connected to VidIQ through a Chrome extension — analyzes three to five competitor channels, surfaces their best-performing topics, writes a 30-second script, and breaks it into scene-by-scene visuals. RankReel turns that script into a finished vertical video with AI voiceover, stock or AI-generated footage, and captions. Then you repeat. “You’re not out here trying to create one viral video,” he says. “You’re creating a system.”

On earnings, he leans on one number: an Air Media Tech analysis of 274 channels putting Shorts RPM at 7 to 20 cents per thousand views, with US-heavy audiences near 33 cents. At 33 cents, he notes, 100 million views is worth roughly $33,000.

The RPM math is real — at the very top of the range

Here’s where a calculator matters. That 33-cent figure isn’t wrong, but it’s the best case, not the base case. Independent 2026 RPM roundups put average Shorts pay closer to $0.03–$0.07 per 1,000 views, with high-value niches (finance, tech, real estate) reaching $0.15–$0.25. Thirty-three cents assumes an almost entirely US, UK, Canada, or Australia audience watching ad-friendly content — a view from a lower-GDP country can be worth 10–20x less.

Run the same 100 million views at average rates and you get $3,000–$7,000, not $33,000. The “interesting facts” niche the video recommends is broad, family-safe, and globally distributed — precisely the profile that pulls RPM toward the low end, not the 33-cent ceiling.

There’s a structural reason the number is soft. YouTube pools all Shorts ad revenue monthly and distributes it by each channel’s share of engaged views, and creators receive 45% of their allocated share. You’re not paid a fixed rate per view; you’re paid a slice of a shared pot that shifts every month with everyone else’s output. More AI-built Shorts flooding the pool doesn’t obviously grow your cut.

Can you even get monetized doing this?

This is the part the video treats as an afterthought, and it’s the whole ballgame.

To earn ad revenue at all, YouTube’s Partner Program requires 1,000 subscribers plus either 4,000 public watch hours in the last 365 days or 10 million public Shorts views in the last 90 days, per YouTube’s own eligibility page. Ten million Shorts views in a rolling 90-day window is not a warm-up target — it’s a real audience. And YouTube has said the bar rises again on February 1, 2027, to 20 million qualified Shorts views in 90 days.

Then there’s the policy that quietly threatens the entire model. In July 2025 YouTube renamed its “repetitious content” rule to the “inauthentic content” policy. Read what it now says: content should “not be mass-produced, generic, repetitive, or manipulative,” and it specifically calls out “AI-generated content made with generic or unoriginal templates giving the impression of mass production without adding the creator’s original, authentic insights.” That language is on YouTube’s monetization policy page.

The video’s method is literally: find what’s already working, make your own version with the same AI voice and stock clips, and repeat. AI editing is fine — YouTube explicitly still allows creative tools. But a channel that’s a template applied at volume, with no original perspective, is exactly what the inauthentic-content rule was rewritten to catch. You can build fifty of these in a weekend and still get denied at review, or demonetized after approval.

Who actually wins this game

Look closely at who’s clearing the money. The channels the video shows already have 180,000 to 320,000 subscribers and years of momentum — YouTube’s recommendation system has learned who to show them to. That’s the moat, and it isn’t AI.

The winners tend to be one of a few types: early movers who scaled a niche before it filled up, operators running many channels at once so a single viral hit pays for the misses, and people who’ve figured out a genuinely distinctive angle (a specific voice, an editing signature, a research depth) that survives the inauthentic-content screen. A brand-new channel copying three competitors is starting at zero against accounts that spent years earning their distribution. The tools are democratized. The audience isn’t.

What you’d realistically earn

For most people trying this cold, the honest first-year range is $0 for months, because you can’t touch ad revenue until you clear 1,000 subscribers and 10 million 90-day Shorts views — and many channels never get there. If you break through, an “interesting facts” channel with a globally mixed audience is likely earning single-digit-to-low-double-digit dollars per million views after the 45% split, not the $330 per million the 33-cent figure implies.

Could a channel hit $33,000 in 90 days? Yes — the ones in the video apparently did. But that’s the visible survivor, not the median attempt, and the video shows you three winners without showing the thousands of AI Shorts channels that posted daily and got demonetized or ignored. That gap between one real peak and the typical result is exactly the kind of unrepresentative earnings claim the FTC moved against in January 2025, when it proposed extending its Business Opportunity Rule to require written substantiation for money-making-opportunity earnings claims. (U.S. readers: that rule targets sellers of opportunities, not YouTubers describing their own results — but the principle, “show your work,” is a good filter for any pitch like this.)

Who this is (and isn’t) for

This makes sense if you can post daily for six to twelve months with no income, you enjoy the format enough to keep going through a dead first quarter, and you’ll add something real — better scripts, original research, a recognizable voice — rather than cloning competitors clip-for-clip. Running several channels to spread the odds helps, as does an audience skewed toward high-RPM countries.

It’s a poor fit if you need money this quarter, if you’re expecting “passive” income (there’s nothing passive about clearing 10 million views), or if your plan is to mass-produce generic templated Shorts — that plan collides directly with YouTube’s rules.

If you want to see how this same “AI runs the content” promise plays out in other formats, we’ve checked similar pitches in Claude Code for a $20,000/month YouTube channel and running an entire content business with AI.

What to remember

The video sells a tidy machine — three tools, one repeatable loop, $39K on the board. The tools work, and a few channels really did earn those sums. What’s missing is the 10-million-view monetization wall, an RPM that’s likely a fraction of the 33 cents quoted, a shared revenue pool you don’t control, and a policy written specifically to demonetize mass-produced AI templates. Do the unspoken work — real distribution, a real angle — and there’s a business here. Skip it, and you’ve built a content machine YouTube won’t pay.